Quoting M H (mh000151@yahoo.com):
> After some googling I found out that the root cause of the recent mount problems with extended partitions (maybe also mounting disks like sda)
> are related to the infinite loop kernel bug in __getblk_slow.
>
> (see the explanation: https://patchwork.kernel.org/patch/1189121/)
> The affected debian kernels should get fixed with the patch offered in kernel 3.2.24 (http://lwn.net/Articles/508225/).


Thanks for the investigation and your care looking out for the root
cause of this.

This bug, and probably all other "hang" bugs, should be reassign to
the kernelpackage.

Before doing so, I'm awaiting for input from a Kerneal Team person.

reassign 684293 linux
severity 684293 critical
thanks

Quoting Brian Potkin (claremont102@gmail.com):
> On Sat 11 Aug 2012 at 19:07:33 +0200, Christian PERRIER wrote:
>
> > Quoting M H (mh000151@yahoo.com):
> > > After some googling I found out that the root cause of the recent mount problems with extended partitions (maybe also mounting disks like sda)
> > > are related to the infinite loop kernel bug in __getblk_slow.
> > >
> > > (see the explanation: https://patchwork.kernel.org/patch/1189121/)
> > > The affected debian kernels should get fixed with the patch offered in kernel 3.2.24 (http://lwn.net/Articles/508225/).
> >
> >
> > Is it me or do both these URL return 404?
>
> Quickly tested both of them a minute ago. They worked for me.


OK, reassigning to the Linux kernel.

To Debian Kernel team: this bug has been revealed in D-I beta1. Any
attempt to mount an extended partition (which is done in both
mountmedia and iso-scan) loops forever and D-I appears to be
"hanged". This happens for instance each time D-I looks for missing
firmware on existing devices.

We had several bug reports about this. mountmedia has been fixed (more
with a workaround than a real fix) but iso-scan is still affected.

So, it seems that the patch described above is highly wished for
wheezy. It would be great to have it before D-I beta2, indeed.

For that reason, I set the bug's severity to critical (breaks
unrelated packages).

reassign 684293 linux
severity 684293 critical
thanks

Quoting Brian Potkin (claremont102@gmail.com):
> On Sat 11 Aug 2012 at 19:07:33 +0200, Christian PERRIER wrote:
>
> > Quoting M H (mh000151@yahoo.com):
> > > After some googling I found out that the root cause of the recent mount problems with extended partitions (maybe also mounting disks like sda)
> > > are related to the infinite loop kernel bug in __getblk_slow.
> > >
> > > (see the explanation: https://patchwork.kernel.org/patch/1189121/)
> > > The affected debian kernels should get fixed with the patch offered in kernel 3.2.24 (http://lwn.net/Articles/508225/).
> >
> >
> > Is it me or do both these URL return 404?
>
> Quickly tested both of them a minute ago. They worked for me.


OK, reassigning to the Linux kernel.

To Debian Kernel team: this bug has been revealed in D-I beta1. Any
attempt to mount an extended partition (which is done in both
mountmedia and iso-scan) loops forever and D-I appears to be
"hanged". This happens for instance each time D-I looks for missing
firmware on existing devices.

We had several bug reports about this. mountmedia has been fixed (more
with a workaround than a real fix) but iso-scan is still affected.

So, it seems that the patch described above is highly wished for
wheezy. It would be great to have it before D-I beta2, indeed.

For that reason, I set the bug's severity to critical (breaks
unrelated packages).

# symptoms appear with v3.2.19~14 (block: don't mark buffers beyond
# end of disk as mapped, 2012-05-11)
reassign 684293 src:linux linux-2.6/3.2.19-1
# fixed by v3.2.24~63 (block: fix infinite loop in __getblk_slow,
# 2012-07-12)
tags 684293 + upstream fixed-upstream
# there's no such thing as a package unrelated to kernel
severity 684293 serious
quit

Hi Christian,

>>> Quoting M H (mh000151@yahoo.com):

>>>> After some googling I found out that the root cause of the recent mount problems with extended partitions (maybe also mounting disks like sda)
>>>> are related to the infinite loop kernel bug in __getblk_slow.
>>>>
>>>> (see the explanation: https://patchwork.kernel.org/patch/1189121/)
>>>> The affected debian kernels should get fixed with the patch offered in kernel 3.2.24 (http://lwn.net/Articles/508225/).

Marking so, thanks.


--
To UNSUBSCRIBE, email to debian-kernel-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Archive: http://lists.debian.org/20120811202849.GB2943@mannheim-rule.local